| If the procedure call to retrieve the data
for the current query fails, the corrected version of
CSV_Database() now switches to the 2nd database from that point
onwards. The basic failover/failback approach used is to use the
primary until such time as it becomes unavailable and at that time,
we swap to the standby database indefinitely. In the event of a
failure accessing the standby database, we swap back to the primary
and so on.
The reason for this ‘bistable’ approach is that it can take a
long time (10s of seconds or minutes in some cases) so it is
worthwhile avoiding connection attempts to a database which was not
operating on the previous call.
Citect has confirmed this to be a problem in CitectHMI/SCADA
versions 5.50 to 6.10SpkB. This problem has been fixed in version
7.0.
|
